The LPS used for treatments in both the in vivo and in vitro models was from Invivogen (#LPS-EK). M-CSF was purchased from StemCell (#78059.1). Antibodies used for flow cytometry were purchased from BD Bioscience: BV650-CD80, PerCP-Cy5.5-Gr1, PE-CD86, PE-CF594-CD40, BV786-CD45, BUV395-CD11c, PerCP-Cy5.5-CD11b, FITC-CD19, BUV496-CD4, BV421-CD206, APC-TLR4, purified CD16/CD32, or BioLegend: PE-Cy7-MHCII, APC fire 750-F4/80, AF647-CD206, zombie aqua. Antibodies used for immunocytochemistry experiments include BV786-CD45 (BD Bioscience #564225), TNF-α (Abcam #ab6671), IL-1β (Abcam #ab205924), IRAP (Cell Signaling #6918) and TLR4 (Abcam #ab13556). DuoSet ELISA kits were purchased from R&D Systems (TNF-α #DY410, IL-1β #DY401, IL-6 #DY406).

WT and global IRAP KO (14) C57BL/6J male and female mice aged 9-15 weeks were used for splenocyte experiments and aged 9-12 weeks were used for BMDM experiments. All experiments were approved by Monash University Animal Ethics Committee and conducted in accordance with the National Health and Medical Research Council’s, Australian code for the care and use of animals for scientific purposes.

To stimulate a systemic inflammatory response, WT and IRAP KO mice were injected (i.p.) with 3 mg/kg LPS or 0.9% v/v saline as the vehicle control once for 4- or 24 hours, or once daily for 2 days (2 injections total; 48-hour endpoint). The body weight of each mouse was recorded before each injection and before culling. Whole blood was collected via cardiac puncture into heparinized tubes and centrifuged at 1,500 rpm for 10 minutes at 4°C. Plasma was stored at -80°C for later analysis. Organs including the spleen, lungs and heart were collected from each animal for further experimentation.

To isolate splenic cells, the spleen from each mouse was dissected, weighed and placed in RPMI 1640 media until being pushed through a 100 µm cell strainer in the presence of RPMI 1640. Collected cells were centrifuged at 1,500 rpm for 5 minutes at room temperature and the resulting pellet was resuspended in ammonium chloride potassium (ACK) lysis buffer (0.15 M NH₄Cl, 0.01 M KHCO₃, 0.0001 M EDTA-Na₂, pH = 7.3, filtered) for 3 minutes to lyse red blood cells. Following lysis, RPMI 1640 was added to adjust the volume of each tube to 10 ml and cells were centrifuged at 1,500 rpm for 5 minutes. The white cell pellet was resuspended in 1 ml of RPMI 1640 and a cell count was conducted using a Countess automated cell counter. A 100 µl suspension containing 1 x 10⁶ splenic cells was prepared for flow cytometry. The remaining cells were frozen at -80°C in RPMI 1640 + 10% DMSO.

To isolate cells from the bone marrow, the tibia and femur were collected from WT and IRAP KO mice. All tissue was carefully removed from the intact bones which were then placed in complete medium consisting of RPMI-1640 supplemented with 10% heat inactivated FBS, 1% penicillin/streptomycin, 2% L-glutamine, 2% HEPES buffer and 0.1 mM 2-mercaptoethanol. The bones were transferred to a biosafety cabinet and appropriate aseptic technique was used for all following steps. The bones were soaked for 1 minute in 80% ethanol and then moved into a petri dish containing complete medium. To expose the bone marrow, the ends of each bone were removed, and bones were placed in a 0.2 ml tube with 3-4 holes at the bottom made using a 19 Ga needle. This tube was placed inside a 1.5 ml tube and subsequently centrifuged at 9,000 rpm for 2 minutes at room temperature to flush out the bone marrow. The resulting bone marrow pellet was resuspended in 200 µl of complete medium and transferred to a 15 ml tube. The suspension was incubated in 3 ml of ACK lysis buffer for 2 minutes to lyse red blood cells. Following addition of complete media to reach a total suspension volume of 10 ml, cells were centrifuged at 1,400 rpm for 4 minutes at room temperature. The pellet of white blood cells was resuspended in 10 ml of complete medium and cells were counted using a Countess automated cell counter. Cells were seeded at a density of 0.5 x 10⁶ cells/ml in 6-well plates in the presence of 10 ng/ml M-CSF in 3 ml of complete medium. For immunocytochemistry experiments, cells were grown on 13 mm coverslips in 6-well plates. Cells were maintained in an incubator at 37°C and 5% CO₂ for 3 days. On day 3, small adherent clusters of BMDM were present. To change media, 1.5 ml of media was removed and replaced with 1.5 ml of fresh complete media with 10 ng/ml of M-CSF. On day 6 of culture, BMDM had a spindle-like morphology and were ready for treatment.

To investigate the role of IRAP in LPS-stimulated macrophage activation, BMDM from both WT and IRAP KO mice were treated for 24 hours with 100 ng/ml LPS on day 6 of culture to stimulate their differentiation into M1 macrophages. Subsets of BMDM were simultaneously treated with IRAP inhibitors (HFI-419 at 10 µM or SJM4-164 at 90 µM) at concentrations 10-fold greater than their IC₅₀ for IRAP.

After 24 hours of treatment, conditioned media from each well were collected for cytokine analysis and BMDM were processed depending on the experiment. For flow cytometry and Western blot experiments, cells were gently scraped into 1.5 ml of ice-cold Dulbecco’s phosphate buffered saline (DPBS) and collected into 1.5 ml tubes. Following centrifugation at 1,400 rpm at 4°C for 4 minutes, cell pellets were either immediately stained for flow cytometry or stored at -80°C for Western blot analysis. For immunocytochemistry experiments, BMDM cultured on 13 mm coverslips were washed briefly with DPBS and then fixed for 15 minutes with 4% PFA. Fixed cells were stored at 4°C.

To stain cells for flow cytometry, 1 x 10⁶ splenocytes or 200 µl of BMDM resuspended in DPBS to a concentration of 0.5 x 10⁶ cells/ml, were transferred to each well of a 96-well conical bottom clear plate. The plate was centrifuged at 1,400 rpm at 4°C for 4 minutes and the resulting supernatant discarded by tipping the plate upside down. Stain mixes (see materials for antibodies used) and fluorescence minus one (FMO) controls were prepared in flow cytometry buffer (DPBS + 2% heat inactivated FBS) and 30 µl was added to the appropriate well. Cells were resuspended well and incubated in the dark for 15 minutes. Cells designated as unstained controls were resuspended in flow cytometry buffer alone. After the 15-minute incubation, 150 µl of flow cytometry buffer was added to each well to flood the stain and the plate was centrifuged at 1,400 rpm at 4°C for 4 minutes. The resulting supernatant was discarded by tipping the plate upside down. Cells were resuspended in 30 µl of viability stain (zombie dye; 1:500 concentration) except for the unstained and FMO controls and incubated for 15 minutes in the dark. Flow cytometry buffer (150 µl) was added to each well and the plate was centrifuged again at 1,400 rpm at 4°C for 4 minutes. The cell pellet was fixed by resuspension in 60 µl of 1% PFA (diluted from 4% PFA with DPBS). Stained cells were transferred into separate labeled flow tubes and stored at 4°C until flow cytometry was conducted.

All flow cytometry data was acquired using Fortessa X20 flow cytometers (BD Biosciences) at Flowcore (Monash University) and analyzed using FlowJo (Tree Star, Ashland, OR). The gating strategy described in Supplementary Figure 4 was used in all splenocyte flow cytometry experiments. Firstly, all cells were gated based on their size and granularity using FSC-A/SSC-A. Single cells were then gated using FSC-A/FSC-H. Live cells were then gated using Zombie/SSC-A where live cells were negative for zombie/viability staining. Immune cell populations including CD45⁺ CD11b⁺ myeloid cells, CD11c⁺ DCs, CD45⁺ F4/80⁺ CD206^- M1 macrophages, CD45⁺ F4/80⁺ CD206⁺ M2 macrophages, CD4⁺ CD45⁺ T helper cells and CD19⁺ CD45⁺ B cells, were subsequently gated. The frequencies of each cell population were analyzed as well as the geometric means of the activation markers CD40, CD80, CD86 and MHCII in all myeloid cell populations.

The gating strategy described in Supplementary Figure 4 was used in all BMDM flow cytometry experiments. Firstly, all cells were gated based on their size and granularity using FSC-A/SSC-A. Single cells were then gated using FSC-A/FSC-H. Live cells were then gated using Zombie/SSC-A where live cells were negative for zombie/viability staining. Immune cell populations including CD45⁺ F4/80⁺ CD206^- M1 macrophages, CD11c⁺ DCs and CD11c^- MHCII⁺ Gr-1⁺ MDSCs, were subsequently gated. It is worth noting that myeloid progenitor cells isolated from the bone marrow and cultured in the presence of M-CSF are expected to consist predominantly (>90%) of CD11b⁺ F4/80⁺ macrophages (15). Therefore, there were very few other immune cell populations examined in this study. The frequencies of each cell population were analyzed as well as the geometric means of the activation markers CD40, CD80, CD86 and MHCII in macrophage populations.

The expression of the pro-inflammatory cytokines TNF-α, IL-1β and IL-6 in the plasma of vehicle- or LPS-treated mice or secreted by BMDM in the conditioned media was measured using mouse DuoSet ELISA kits (R&D Systems). ELISAs were run according to the manufacturer’s instructions.

To visualize the expression of various cytokines in CD45⁺ BMDM, immunocytochemistry was conducted. Cells fixed with 4% PFA were washed 3 times for 10 minutes each with 1x PBS. Coverslips were removed from the 6-well plates and placed on a microscope slide, with the cells facing upwards. Cells were blocked and permeabilized with 10% normal goat serum/0.2% triton X-100 for 1 hour, followed by overnight incubation at room temperature with the anti-CD45, anti-TNFα, anti-IL-1β, anti-IL-6, anti-IRAP or anti-TLR4 antibody diluted in antibody diluent (all 1:500 dilution). For dual labeling, the cells were incubated with both primary antibodies overnight. The following day, the primary antibody was removed and the cells were rinsed with PBS in a series of three washes (10 minutes each). The cells were subsequently incubated with the appropriate secondary antibody for 2 hours at room temperature (1:500 dilution) and the three washes with PBS were repeated. One drop of Vectashield mounting medium containing DAPI was placed onto each coverslip and a rectangular coverslip was secured on top. After ensuring no bubbles were present, the mounting medium was left to dry for 10 minutes before imaging.

All imaging was performed using a Zeiss Axio Imager M1. At least 6-8 images at 20x magnification were taken per treatment group, with each field of view containing cells at approximately the same level of confluency. An ImageJ macro was developed to quantify the positive immunofluorescent staining area of TLR4 in a field of view. This number was normalized to the total cell count for the same image, averaged for each group and presented relative to the mean of the male WT control group. The cell counts, which used thresholded-DAPI stained nuclei, were completely automated as part of the ImageJ macro.

To quantify the colocalization of two proteins such as IRAP and TLR4, an ImageJ macro was written to measure the intersection using the “AND” option in the image calculator tool in a given field of view. This number was then normalized to the total positive staining area in the same image, calculated as the combined area measured using the “Add” option in the image calculator tool minus the intersection, to give percent area of colocalization. The percent area colocalization was finally averaged for each group.

To measure the protein expression of IRAP in BMDM from WT and IRAP KO mice, standard Western blotting was conducted. To extract the protein, BMDM cell pellets stored at -80°C were thawed and lysed in 40 µl of buffer consisting of 0.5% triton X-100 and 1% protease/phosphatase inhibitor cocktail in 50 mM Tris/150 mM NaCl. Following a 30-minute incubation on ice, samples were centrifuged (6,000 rpm for 10 minutes at 4°C) and the protein-containing supernatants were transferred to new tubes.

All experiments were conducted under non-reducing conditions where lysates were diluted in 1.5x laemmli buffer (without 2-mercaptoethanol) and heated at 70°C for 5 minutes. For BMDM lysates, ~80-100 µg of protein was loaded per well (calculated retrospectively by measurement of the protein concentration of diluted lysates using a nanodrop). Before loading of samples, 10% gels (15 wells, 1mm thickness) were prepared using a TGX Stain-free FastCast Acrylamide starter kit and standard Western blotting protocol was followed whereby protein separated in the gel is transferred to a membrane, blocked with 5% skim milk for 1 hour, incubated with anti-IRAP antibody overnight at 4°C and the appropriate HRP-conjugated secondary antibody the following day. Individual protein bands were quantified by measuring the optical density (OD) per unit area using ImageLab software. Correction for differences in protein per well was normalized to the protein concentration measured using a nanodrop. All protein expression is presented as a relative ratio to the mean of the male WT control group to allow for examination of fold changes.

Results are presented as mean ± SEM and statistical analysis was conducted using GraphPad Prism software. In the in vivo LPS model, two-way ANOVAs with Tukey’s multiple comparisons were used to investigate the effect of LPS or genotype on the body weights of mice, the frequency of different cell populations and the expression of different activation markers. The main effects, namely treatment (vehicle vs LPS) and genotype (WT vs KO) as well as the interaction between the two factors, generated from the two-way ANOVAs, are presented in Supplementary Table 1 whereas results from the post-hoc Tukey’s multiple comparison are presented in the figures in the main results text. It is worth noting there is some variability in the basal frequencies of some of the immune cell populations at different timepoints, potentially due to them being conducted on different days. To compare the spleen weights of all groups, a three-way ANOVA with Tukey’s post-hoc test was used with LPS treatment, genotype and sex as the independent variables. To determine if there was a relationship between the change in body weight and spleen weight, correlational analysis with Pearson’s correlation coefficient was conducted using a two-way ANOVA with Tukey’s post-hoc analysis. The quantification of cytokine expression in the plasma of vehicle- and LPS-treated mice, measured using sandwich ELISAs, was analyzed by two-way ANOVAs with Tukey’s post-hoc test with LPS treatment and genotype as the independent variables.

In the in vitro BMDM model, the frequency of different cell populations and the expression of activation markers across groups were analyzed using a two-way ANOVA with Tukey’s post-hoc analysis with the LPS treatment and genotype as independent variables. The quantification of cytokine expression in the conditioned media of BMDM, measured using sandwich ELISAs, was analyzed by three-way ANOVAs with Tukey’s post-hoc test with LPS treatment, genotype and sex as the independent variables. In Western blot experiments, a three-way ANOVA with Tukey’s post-hoc test was used to compare IRAP protein expression across groups in BMDM with independent variables of LPS treatment, genotype and sex. A two-way ANOVA with Tukey’s post-hoc was used to compare IRAP and CD45 colocalization in immunocytochemistry experiments. To compare the expression of TLR4 in BMDM, colocalization of IRAP or TLR4 in BMDM and expression of TLR4 at the cell surface in M1 macrophages in the spleen, three-way ANOVAs with Sidaks’s post-hoc analysis were conducted.

There were no notable genotype or sex differences in body weight, spleen weight or clinical severity scores after a single intraperitoneal dose of LPS for either 4- or 24-hours, or two daily doses over 48-hours. LPS treatment induced a significant reduction in body weight across both genotypes at each timepoint, with a mean loss of 2.4 ± 0.3% at 4 hours (p = 0.0126), 10.8 ± 0.8% at 24 hours (p < 0.0001) and 15.8 ± 0.6% at 48 hours (p < 0.0001, vs baseline, two-way ANOVAs; Figure 1A ). LPS-treated mice all displayed a greater clinical severity score, with decreased locomotor activity and lack of response to handling, similar to previous reports in this sepsis model (13, 16). Significant time-dependent, progressive increases in the spleen weight relative to the body weight were recorded in all LPS-treated groups compared to vehicle controls after 24-hours (p = 0.1360 (4hr), p < 0.0001 (24hr), p < 0.0001 (48hr), three-way ANOVAs; Figure 1B ), regardless of genotype. There was also a significant inverse correlation between the body weight change and spleen weight 24 hours after LPS treatment, with mice that lost more body weight having larger spleens (R² = 0.14, p = 0.23 (4hr), R² = 0.39, p = 0.03 (24hr), R² = 0.55, p<0.0001 (48hr); Figure 1C ). Together, these findings validate the systemic inflammation model, confirming that LPS treatment causes significant weight loss as well as inducing an increase in immune cell infiltration into the spleen.

DCs participate in all stages of gram-negative sepsis initiation and progression, with documented alterations in their frequency, differentiation and activation state (17). IRAP is highly expressed in DCs where it has been shown to play a number of physiological roles (7, 11, 18). The effect of global gene deletion of IRAP on the acute inflammatory response to LPS treatment was investigated by comparing changes in splenic DC frequency and activation, via the markers CD40, CD80, CD86 and MHCII, in WT and IRAP KO, male and female mice. In control male mice, 2.9 ± 0.6% of cells in the spleen were CD11c⁺ DCs, while in control female mice, 3.4 ± 0.9% of cells were DCs. These frequencies are similar to those reported in previous studies in C57BL/6 mice (19, 20). Supplementary Table 1 summarizes the analysis from the two-way ANOVA showing either an overall treatment (vehicle vs LPS) or genotype (WT vs IRAP KO) effect on the distinct immune cell frequencies or the activation markers. A significant LPS treatment effect on the frequency of the DCs was detected in both sexes at 24 hours ( Supplementary Table 1 ). Post-hoc analysis revealed that this significant decrease in DC frequency was only observed in the WT group in the female mice (p = 0.0039, two-way ANOVA Tukey’s post-hoc test; Supplementary Figure 1B ). Furthermore, in female mice, there were significantly greater fold-increases in CD40, CD86 and MHCII markers in DCs following LPS stimulation at 24 hours ( Supplementary Table 1 ; Supplementary Figure 1C ) and by 48h, the most of these activation markers were normalized to levels closer to or even lower than that of vehicle controls. Overall, the only significant genotype effect in DCs were on the activation markers in cells isolated from spleens of female mice 24 hours after LPS stimulation ( Supplementary Table 1 ). This suggests a sex-dependent role for IRAP in moderating the DC response to LPS induced inflammation.

Macrophages have diverse roles in all phases of gram-negative sepsis as part of both the innate and adaptive immune response (21). Unlike other immune cell types, they also have the unique ability to become polarized into a pro-inflammatory M1 phenotype or anti-inflammatory M2 phenotype depending on cues in the microenvironment (21). LPS is a potent stimulator of M1 macrophage activation in vitro (22). While the categorization of M1 and M2 macrophages is important experimentally to gain insights into their functional roles, the expressions of many macrophage markers are continually changing suggesting that macrophage polarization is a dynamic process.

M1 macrophages were defined as CD45⁺ monocytes that expressed F4/80 and lacked the M2 macrophage marker, CD206 ( Figures 2A , 3A ). In male mice, LPS treatment significantly increased the frequency of M1 macrophages in the spleen at 24 and 48 hours from a mean across genotypes of 0.91 ± 0.007% to 1.64 ± 0.09% and 5.55 ± 0.29% to 8.47 ± 0.06%, respectively ( Supplementary Table 1 ; Figure 2B ). In females, the increase in M1 frequency was only significant at 48 hours post LPS treatment from 4.98 ± 0.21% to 9.20 ± 0.24% ( Supplementary Table 1 ; Figure 3B ). No significant genotype effect was detected in the frequency of M1 macrophages or expression of activation markers CD40, CD80, CD86 or MHCII in either sex ( Supplementary Table 1 ). Further examination of the activation markers using post-hoc analysis revealed a marginal increase in CD40 expression in M1 macrophages from the male IRAP KO mice (p = 0.0124, two-way ANOVA Tukey’s post-hoc test) ( Figure 2C ). Interestingly, evaluation of the frequency curves revealed that in male IRAP KO mice, there appears to be a higher frequency of M1 macrophages expressing MHCII at 24 hours and more so at 48 hours after LPS treatment ( Figure 2D ), an effect that was not present in female mice ( Figures 3A, C, D ).

Activated M2 macrophages were defined as CD45⁺ monocytes expressing F4/80 and CD206 ( Figures 4A , 5A ). The frequency of the M2 macrophages and their expression of activation markers after an LPS challenge were evaluated. The earliest significant changes in M2 macrophages were detected at the 24-hour timepoint where a transient 2-3-fold increase in cell frequency in response to LPS was followed by 4-5-fold decrease at 48 hours ( Figures 4B , 5B ). This suggests that the environment encountered by infiltrating monocytes in the spleen promoted polarization to an M2 anti-inflammatory phenotype at 24 hours which transformed to the M1 pro-inflammatory phenotype at 48 hours. This trend was consistent across males and females ( Figures 4B , 5B ). However, in male WT mice, an exaggerated increase in M2 frequency in response to LPS was detected at 24 hours compared to the IRAP KO. This pattern was completely reversed at 48 hours after LPS treatment where the frequency was significantly decreased in both genotypes and sexes ( Figures 4B , 5B ). As for the expression of the activation markers, IRAP gene deletion significantly dampened the LPS-induced increases in CD40 at the 24-hour timepoint but only in male mice ( Figures 4C , 5C ).

Gram-negative sepsis results in dysregulation and deficits in the adaptive immune response, with depletion of functional T and B cells due to increased cell death or exhaustion (23). IRAP is known to participate in specific components of the adaptive immune response in T cells (10), although there has been no investigation into IRAP in B cells. In the spleens of vehicle-treated mice across both genotypes, 8.79 ± 1.02% of cells were CD4⁺ CD45⁺ helper T cells in males ( Supplementary Figure 2B ) and 11.21 ± 1.34% of cells were helper T cells in females ( Supplementary Figure 2C ). There was also a very high proportion of CD19⁺ CD45⁺ B cells in the spleens of control mice across genotypes, with a frequency of 35.26 ± 2.11% in males and 37.21 ± 2.50% in females, similar to previous reports (20). There were significant decreases in helper T cell frequency 24 and 48 hours following LPS treatment in both males and females ( Supplementary Table 1 ). This supports reports on T cell depletion in sepsis (24). Notably, there was also a significant genotype effect in T cell frequency at all timepoints in female mice ( Supplementary Table 1 , Supplementary Figure 2C ). LPS administration induced significant increases in B cell frequency in both sexes at 48 hours ( Supplementary Table 1 ) but no genotype effect was observed.

The altered secretion of cytokines from cells such as macrophages is a key characteristic of gram-negative sepsis with an uncontrolled overproduction of pro-inflammatory mediators in acute phases (3). There were overall increases in pro-inflammatory tumor necrosis factor (TNF)-α, interleukin (IL)-1β and IL-6 in the plasma of LPS-treated compared with vehicle-treated mice at the 4-hour timepoint ( Supplementary Figure 3 ), although only IL-1β and IL-6 were statistically significant in both males (IL-1β: p=0.0224, IL-6: p=0.0191; two-way ANOVAs) and females (IL-1β: p=0.0020, IL-6: p=0.0058; two-way ANOVAs). There were no significant genotype effects in either sex. It is worth noting that while essentially none of the cytokines were detected in any plasma samples from control mice, LPS induced 10-to-100-fold increases in protein levels of TNFα and IL-1β and 1000-fold increases in protein levels of IL-6. Pro-inflammatory cytokines levels were subsequently restored to undetectable control levels at the 24- and 48-hour timepoints following LPS stimulation. This observation is supported by a previous study where cytokine production occurred within 2 hours of LPS injection with peak increases in circulating TNF-α and IL-6 detected at this time point before returning to baseline over the following 6 hours (25).

Given the complexity of the acute inflammatory response in vivo and the plethora of factors that could contribute to the regulation of macrophage activation, it is difficult to ascertain whether locally produced IRAP within the macrophages themselves plays a role in LPS-responsiveness and activation state. To explore this concept under controlled conditions in vitro, bone-marrow derived macrophages (BMDM) were treated with LPS to stimulate polarization into an M1 pro-inflammatory phenotype (26, 27).

Examination of BMDM from male WT mice revealed that LPS stimulation increases the proportion of M1 macrophages from 4.98 ± 1.24% to 16.34 ± 4.04% (p = 0.3019, two-way ANOVA; Figure 6A, B ). Interestingly, this response was exacerbated in cells isolated from male IRAP KO mice with an increase from 8.27 ± 2.91% to 34.03 ± 6.80% (p=0.0029, two-way ANOVA; Figure 6B ). This heightened response in cells deficient for IRAP was mirrored in the expression of the activation marker, CD40, following LPS treatment with a significant genotype (p = 0.0102) and treatment effect (p = 0.0089) as well as a significant interaction between the two factors (p=0.0102, two-way ANOVA; Figure 6D ). There were inconsistent, non-significant LPS-mediated effects on the expression of CD80, CD86 and MHCII ( Figures 6C, D ). Notably, treatment with small molecule IRAP inhibitors developed by our laboratory, HFI-419 (28) and SJM4-164 (unpublished), did not exert any consistent effects on any of the measured variables (data not shown), supporting previous findings of the lack of the catalytic site involvement in regulating the immune response (7, 10). Overall, IRAP deficiency resulted in an increased frequency of M1 macrophages and expression of CD40 suggesting that the enzyme plays a role in dampening M1 macrophage activation thus controlling the pro-inflammatory response to an acute immune challenge.

Unlike the macrophages derived from male mice, LPS-induced significant and consistent increases in the frequency of M1 macrophages, as well as in all the activation markers, of female mice ( Figure 7 ). However, IRAP gene deletion resulted in a significant suppression of the LPS response on these markers in the BMDM ( Figure 7C ).

Furthermore, sandwich ELISAs conducted on conditioned media from BMDMs revealed significant increases in the secretion of the pro-inflammatory cytokines, TNF-α and IL-1β with LPS treatment from cells derived from male WT and IRAP KO mice ( Figure 8 ) with a heightened TNF-α and IL-1β secreted in the IRAP deficient BMDM. This heightened response was not observed in cells derived from female mice. These results provide a clearer picture of the role of IRAP on the production and secretion of pro-inflammatory cytokines from macrophages in response to LPS stimulation.

To further examine the physiological role of IRAP in macrophages, the expression and distribution of the enzyme was examined in BMDM. LPS treatment induced a 3-4-fold increase in expression of IRAP protein in BMDM lysates from both male and female WT mice ( Figures 9A, B ). This increase in IRAP expression with LPS is supported by previous findings in mouse peritoneal macrophages (29). Immunostaining of IRAP in CD45⁺ BMDM revealed that the enzyme is located at the plasma membrane as well as in endosomal compartments. However, the changes in the subcellular distribution of IRAP in response to LPS stimulation is difficult to deduce as the activation of these BMDM resulted in an increase in cell size and shape ( Figure 9C ). Across all groups, there was only 12.85 ± 0.85% colocalization of IRAP and CD45, validating that IRAP is not highly expressed at the plasma membrane but largely remains confined to endocytic vesicles ( Figure 9D ). This semi-quantitative result supports previous observations that ~10% of IRAP is at the cell surface in a basal state at a given point in time (9).

The N-terminal domain of IRAP has been found to regulate the trafficking of specialized endosomes containing pro-inflammatory cargo in DCs (7), consequently restricting downstream pro-inflammatory pathways. It is possible a similar mechanism exists in macrophages, whereby IRAP may regulate the trafficking of components associated with a pro-inflammatory response such as the LPS receptor, TLR4, and thus control the extent of its downstream signaling.

Immunofluorescent staining revealed TLR4 is not highly expressed in basal CD45⁺ BMDM and has a dispersed cytosolic distribution ( Figure 10A ). A significant change in TLR4 expression occurred with LPS stimulation with ~3-fold increases in male-derived cells and only very minor, inconsistent increases in female-derived cells ( Figure 10B ). This supports previous literature which demonstrate increases in TLR4 expression with LPS stimulation (30, 31). Surprisingly, there was also a significant sex effect with females expressing lower levels of TLR4 than males both at baseline and after LPS challenge (p=0.002, three-way ANOVA). There was no effect of genotype on TLR4 expression. This suggests that IRAP does not influence the expression of TLR4 at steady state or in response to LPS.

IRAP appeared to influence the subcellular distribution of TLR4 in a basal state in male-derived macrophages where comparison of the percentage of TLR4 colocalized with the cell surface marker, CD45, revealed that cells derived from male IRAP KO mice appear to have a higher basal cell surface level of TLR4 (17.28 ± 5.49%) compared to cells derived from WT mice (9.45 ± 0.24%; Figure 10C ). This result was recapitulated in vivo, with higher basal cell surface levels of TLR4 in CD45⁺ F4/80⁺ CD206^- M1 macrophages in the spleen of male IRAP KO mice (11.50 ± 1.25 MFI/geometric mean of expression) compared with male WT mice (6.66 ± 1.83 MFI/geometric mean of expression), measured via flow cytometry ( Figure 10D ). In contrast, there was a similar cell surface expression of TLR4 in steady state cells from female mice across genotypes. This difference in the cellular localization of TLR4 in a basal state that may be regulated by IRAP could dictate the responsiveness of the cell to LPS.